Search

Coral reefs are biodiverse ecosystems that rely on trophodynamic transfers from primary producers to consumers through the detrital pathway. The sponge loop hypothesis proposes that sponges consume dissolved organic carbon (DOC) and produce large quantities of detritus on coral reefs, with this turn-over approaching the daily gross primary production of the reef ecosystem. In this study, we collected samples of detritus in the epilithic algal matrix (EAM) and samples from potential sources of detritus over two seasons from the forereef at Carrie Bow Cay, Belize. We chose this location to maximize the likelihood of finding support for the sponge loop hypothesis because Caribbean reefs have higher sponge abundances than other tropical reefs worldwide and the Mesoamerican barrier reef is an archetypal coral reef ecosystem. We used stable isotope analyses and eDNA metabarcoding to determine the composition of the detritus. We determined that the EAM detritus was derived from a variety of benthic and pelagic sources, with primary producers (micro- and macroalgae) as major contributors and metazoans (Arthropoda, Porifera, Cnidaria, Mollusca) as minor contributors. None of the sponge species that reportedly produce detritus were present in EAM detritus. The cnidarian signature in EAM detritus was dominated by octocorals, with a scarcity of hard corals. The composition of detritus also varied seasonally. The negligible contribution of sponges to reef detritus contrasts with the detrital pathway originally proposed in the sponge loop hypothesis. The findings indicate a mix of pelagic and benthic sources in the calmer summer and primarily benthic sources in the more turbulent spring.

Abstract Climate change is a critical threat to northern freshwater ecosystems, yet many remote areas are data deficient in terms of biodiversity information. Generating community composition data through collection of environmental DNA (eDNA) is less labor‐intensive than traditional sampling methods and is being increasingly used in areas that have been historically difficult to sample such as northern freshwater habitats. Here, we employed eDNA metabarcoding using three mitochondrial markers at 174 coastal river sites, sampled over three years (2019–2021) across a broad region in northeastern North America, Newfoundland and Labrador. We characterized current riverine fish community composition, compared it to traditional sampling records, and quantified the influence of climate on variation in fish community composition. The analysis detected 33 fish species across the region (1–13 per location), including three non‐native species, as well as several new possible range expansions. Variance partitioning with redundancy analysis indicated ~56% of the variation in community composition could be explained by spatial and climate factors (~21% and ~7%, respectively, with an additional ~28% shared). A temporal comparison across a subset of locations with both eDNA and historical records (1965–1985) revealed that more species were detected on average with eDNA sampling, and that sampling method explained a small portion of the variation (~4%) in comparison with space (~10%) and climate (~7%). Ultimately, this work is the most complete survey of freshwater and diadromous fishes present in Newfoundland and Labrador to date, highlights new detections of non‐native species including previously unknown diversity for the region, and provides future direction for the application of eDNA analysis in northern riverine habitats.